Computer system architecture: classification and definition

Modern computer solutions can be classified, based on their attribution to a particular architecture. But what can it be? What are the main approaches to understanding this term?

The architecture of computer systems as a set of hardware components

What is the essence of the concept of "computer system architecture"? By the appropriate term, first of all, one can understand a set of electronic components, of which a PC consists, interacting within a certain algorithm using different types of interfaces.

The main components that are part of the computer system:

• input device;
• Main computing chipset;
• Devices for data storage;
• Components for outputting information.

In turn, each of the marked components can include a large number of separate devices. For example, the main computing chipset can include a processor, a chipset on the motherboard, a graphics processing module. In this case, the same processor can consist of other components: for example, the core, cache memory, registers.

Proceeding, in fact, from the structure of specific hardware components of the PC, it is determined which architecture of the computer system is built. Let's consider the basic criteria according to which those or other computing solutions can be classified.

Classification of computer systems

In accordance with the approach common in the expert environment, computer systems in their architecture may include:

• To the mainframe;
• To a minicomputer;
• To personal computers.

It should be noted that this classification of computing solutions, according to which the architecture of the computer system can be determined, is recognized by many experts as outdated. In particular, the same personal computers today can be divided into a large number of varieties, very dissimilar in purpose and characteristics.

Thus, as computer systems develop, the computer architecture can be classified using changing criteria. Nevertheless, this scheme is considered to be traditional. It will be useful to consider it in more detail. In accordance with it, the first type of computer - those that belong to the architecture of large machines.

Large computers

Large computers, or mainframes, are most often used in industry - as data centers for various production processes. They can be installed powerful, extremely high-performance chips.

The architecture of the computer system under consideration can realize up to several tens of billions of calculations per second. The big computers stand incomparably more expensive than other systems. As a rule, their maintenance requires the participation of a fairly large number of people who have the necessary qualifications. In many cases, their work is carried out within the units organized as a computer center of the enterprise.

Mini-computer

The architecture of computer systems and computer networks on their basis can be represented by solutions classified as a mini-computer. In general, their purpose can be the same as in the case of mainframes: it is very common to use the appropriate type of computers in the industry. But, as a rule, their use is typical for relatively small enterprises, medium-sized businesses, scientific organizations.

Modern mini-computers: opportunities

In many cases, the use of these computers is carried out just to effectively manage intranet networks. Thus, the considered solutions can be used, in particular, as high-performance servers. They are also equipped with very powerful processors, such as, for example, Xeon Phi from Intel. This chip can work at a speed of more than 1 teraflops. The corresponding processor is designed for production on a 22 nm process technology and has a memory bandwidth of 240 GB / s5.

Personal computers

The next type of computer architecture is a PC. It is probably the most common. PCs are not as powerful and high-performance as mainframes and micro-computers, but in many cases are capable of solving tasks in the field of industry and science, not to mention typical user tasks such as launching applications and games.

Another remarkable feature that characterizes personal computers is that their resources can be combined. The computing power of a sufficiently large number of PCs can thus be comparable to the performance of higher-end computer architectures, but, of course, reaching their levels nominally with a PC is very problematic.

Nevertheless, the architecture of computer systems and networks based on personal computers is characterized by the universality, in terms of implementation in various industries, accessibility and scalability.

Personal computers: classification

As we noted above, PCs can be classified into a large number of varieties. Among them: desktops, laptops, tablets, PDAs, smartphones - combining PCs and phones.

As a rule, the most powerful and productive architectures are desktops; The least powerful - smart phones and tablets due to the small size and the need to significantly reduce the resources of hardware components. But many of the corresponding devices, especially the top models, in terms of speed, in principle, are comparable with the leading models of laptops and budget desktops.

The marked classification of PCs indicates their universality: in certain varieties they can solve typical user tasks, production, scientific, laboratory. Software, the architecture of computer systems of the appropriate type is in many cases adapted to the use by an ordinary citizen who does not have the special training that a person working with a mainframe or a minicomputer may need.

How to determine the assignment of a computational solution to a PC?

The main criterion for assigning a computational solution to a PC is the fact of its personal orientation. That is, the corresponding type of computer is designed primarily for the use of one user. However, many of the infrastructural resources to which he addresses are of undeniably social nature: this can be traced to the example of using the Internet. While the computing solution is personal, the practical effectiveness in its use can be recorded only if a person gets access to data sources formed by other people.

Classification of software for computer architectures: mainframes and minicomputers

Along with the classification of computers considered above, there are also criteria for classifying programs that are installed on the appropriate types of computers. As for mainframes and their close ones by purpose, and in some cases also by the performance of a mini-computer, they, as a rule, are realized the possibility of using several operating systems adapted for solving specific production tasks. In particular, these OS data can be adapted to launch various automation tools, virtualization, the implementation of industry standards, integration with various types of software applications.

Software classification: personal computers

Programs for ordinary PCs can be presented in versions that are optimized to solve, in turn, user tasks, as well as those that do not require the level of performance that characterizes mainframes and minicomputers. There are, therefore, programs for PC industrial, scientific, laboratory. Software, the architecture of computer systems of the appropriate type depends on the specific industry in which they are applied, from the prospective level of the user's skill: it is obvious that professional solutions for industrial design may not be designed for a person who has only basic knowledge of the use of computer programs.

Programs for PCs in one form or another have in many cases an intuitive interface, various reference documentation. In turn, mainframe and mini-mainframe capacities can be fully used, provided not only following the instructions, but also with regular changes made by the user to the structure of the launched programs: for this, additional knowledge may be needed, for example, related to the use of languages Programming.

Levels of PC software architecture

The concept of "architecture of computer systems" textbook of informatics, depending on the views of its author, can be interpreted in different ways. Another common interpretation of the corresponding term implies its correlation with the levels of the software. At the same time, it does not matter what kind of computer system the corresponding software levels are implemented in.

In accordance with this approach, the computer architecture should be understood as a set of different types of data, operations, characteristics of the software used to maintain the functioning of the hardware components of the computer, as well as creating the conditions under which the user is able to apply these resources in practice.

Software Level Architectures

Experts identify the following main architectures of computer systems in the context of the approach to understanding the corresponding term:

• Digital logical architecture of the computing solution - in fact, the hardware of the PC in the form of various modules, cells, registers - for example, located in the structure of the processor;
• Microarchitecture at the level of interpretation of various microprograms;
• Architecture of translation of special commands - at the assembler level;
• Architecture of the interpretation of the relevant commands and their implementation in the program code, understandable to the operating system;
• The architecture of compilation, allowing to make changes in program codes of those or other kinds of software;
• Architecture of high-level languages, allowing to adapt program codes to the solution of specific user tasks.

The value of the classification of software architecture

Of course, this classification in the context of treating this term as appropriate to software levels can be very conditional. The computer architecture and the design of computer systems, depending on their adaptability and designation, may require other approaches of developers in the classification of software levels, and also, in fact, to an understanding of the essence of the term in question.

Despite the fact that these representations are theoretical, their adequate understanding is of great importance, since it contributes to the development of more effective conceptual approaches to building certain types of computing infrastructure, allows developers to optimize their solutions to the needs of users who solve specific problems.

Summary

So, we have defined the essence of the term "architecture of a computer system", how it can be viewed depending on a particular context. In accordance with one of the traditional definitions, under the appropriate architecture can be understood the hardware structure of the PC, predetermining the level of its performance, specialization, requirements for the qualification of users. This approach involves the classification of modern computer architectures into three main categories - mainframes, minicomputers, and PCs (which, in turn, can also be represented by various types of computing solutions).

As a rule, each type of these architectures is designed to solve certain tasks. Mainframes and minicomputers are most often used in industry. With the help of a PC it is also possible to solve a wide range of production tasks, carry out engineering development - for this, the corresponding architecture of computer systems is also adapted. Laboratory work, scientific experiments with this technique become clearer and more effective.

Another interpretation of the term in question implies its correlation with specific levels of software. In this sense, the architecture of computer systems is a work program that ensures the functioning of the PC, and also creates the conditions for using its computing power in practice to solve specific user tasks.